We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

/pdf/going-deeper-with-convolutions-1yobw2o2ds.pdf

Going deeper with convolutions

Co-Planar Stereotaxic Atlas of the Human Brain—3-Dimensional Proportional System: An Approach to Cerebral Imaging, J. Talairach, P. Tournoux. Georg Thieme Verlag, New York (1988), 122 pp., 130 figs. DM 268

I have developed "tennis elbow" from lugging this book around the past four weeks, but it is worth the pain, the effort, and the aspirin. It is also worth the (relatively speaking) bargain price. Including appendixes, this book contains 894 pages of text. The entire panorama of the neural sciences is surveyed and examined, and it is comprehensive in its scope, from genomes to social behaviors. The editors explicitly state that the book is designed as "an introductory text for students of biology, behavior, and medicine," but it is hard to imagine any audience, interested in any fragment of neuroscience at any level of sophistication, that would not enjoy this book. The editors have done a masterful job of weaving together the biologic, the behavioral, and the clinical sciences into a single tapestry in which everyone from the molecular biologist to the practicing psychiatrist can find and appreciate his or

Principles of Neural Science

http://www.cnbc.pt/jpmatos/29.%20Bradley.pdf?origin=publicationDetail

Measuring emotion: The self-assessment manikin and the semantic differential

Well-being : the foundations of hedonic psychology

The effects of an emotional stimulus prepulse on probe startle response were examined here. Pleasant, neutral and unpleasant pictures were viewed for 6 s, and an acoustic startle probe was presented either 300, 800, 1,300, or 3,800 ms after slide onset, or 300 or 3,800 ms after slide offset. Blink magnitude and onset latency demonstrated (a) an early (prepulse) inhibition effect in which reflexes elicited immediately after slide onset were smaller than reflexes elicited later in the viewing interval, and (b) affective modulation, in which unpleasant stimuli prompted larger reflexes than pleasant. Interactive effects of probe time and picture valence indicated attention/arousal effects early and pleasantness effects late in the picture interval. Effects of both attention and emotion can be simultaneously measured using this startle-probe paradigm, encouraging its use in both basic and clinical contexts.

Pictures as prepulse: attention and emotion in startle modification

A repetition paradigm was used to assess the nature of affective modulation of early and late components of the event-related potential (ERP) during picture viewing. High-density ERPs were measured while participants passively viewed affective or neutral pictures that were repeated up to 90 times each. Both ERP components were modulated by emotional arousal, with ERPs elicited when viewing pleasant and unpleasant pictures different than when viewing neutral pictures. On the other hand, repetition had different effects on these two components. The early occipitotemporal component (150--300 msec) primarily showed a decrease in amplitude within a block of repetitions that did not differ as a function of picture content. The late centroparietal component (300--600 msec) showed a decrease both between and within blocks of repetitions, with neutral pictures eliciting no late positive potential in the final block of the study. The data suggest that the early ERP primarily reflects obligatory perceptual processing that is facilitated by active short-term memory representations, whereas the late ERP reflects increased resource allocation due to the motivational relevance of affective cues.

https://csea.phhp.ufl.edu/Publications/Old/Codispoti2007JCN.pdf

Repetition and Event-related Potentials: Distinguishing Early and Late Processes in Affective Picture Perception

Brain activation by disgust-inducing pictures in obsessive-compulsive disorder.

https://csea.phhp.ufl.edu/Publications/Old/Codispoti,%202006.pdf

Margaret M. Bradley

Papers

Pictures as prepulse: attention and emotion in startle modification

Repetition and Event-related Potentials: Distinguishing Early and Late Processes in Affective Picture Perception

Brain activation by disgust-inducing pictures in obsessive-compulsive disorder.

Repetitive picture processing: autonomic and cortical correlates.

Affect and the startle reflex.